Temporal variation in the results of soil phosphate analyses

Abstract

Soil phosphate (P) values were determined for 49 improved pasture sites on 11 occasions over a 3-year period. Each sample was taken from under an improved pasture on the Central Tablelands of New South Wales and analysed using the Bray No. 1 and Colwell (modified Olsen) tests. Variations in soil P values between samplings over time were significant (P<0.05). For individual sites, the 95% confidence limit, as a percentage of the mean, averaged ¦ 19% for Bray P values and ¦ 13% for Colwell P values. The pattern of variation in P values over time was not significantly (P<0.05) affected by soil P level, soil type or soil test. Variation in P values over time with both tests was significantly (P<0.05) correlated with a general estimate of soil moisture and thermal index for the sampling month. Both Colwell P and Bray P values showed negative correlations with increasing soil moisture or increasing thermal index. The correlation between Colwell P and Bray P values on any one soil type was not reliable enough to allow prediction of one soil-test P value from the other.